Image registration is a mathematical process whereby spatially corresponding points in different images are identified. Simply stated, registration between a pair of images enables one to declare, for example, that: point A in image 1 corresponds to point B in image 2. Solution to the problem of image registration is extremely important in medical imaging research, where increasingly large amounts of imaging data are collected from research subjects using highly sophisticated medical imaging instruments. The principal investigator of the current project has developed a package, the Automatic Registration Toolbox (ART),that allows solving various image registration problems that arise in brain imaging research. However, ART has heretofore been underutilized due to lack of resources for its proper enhancement, documentation, dissemination, and subsequent troubleshooting. Recognizing this problem in general, the National Institutes of Health Neuroscience Initiative recently awarded a contract for the development of a Neuroinformatics Tools and Resources Clearinghouse (NITRC), in order to promote the enhancement, adoptation, distribution, and evolution of neuroimaging informatics tools and resources, such as ART. The broad aim of the present proposal is to take advantage of the opportunity provided by the NITRC initiative, in order to transform ART into a well-documented and user-friendly package that can be easily and seamlessly distributed to the wider research community. Specifically, this project aims to: (1) design a unified and portable graphical user interface for ART; (2) enhance the software's ability to read and write in various image formats; (3) enhance accessibility to the software both for developers, by proving command-line versions of ART modules and linkable software libraries, and for general users through the NITRC; and (4) develop comprehensive and detailed documentation for ART in a variety of electronic formats. [unreadable] [unreadable] [unreadable] [unreadable]